diag.RBS: Diagnostic Analysis - Local Influnce

In santosneto/RBS: Regression models for Birnbaum-Saunders distributions

Description

Diagnostics for the RBS model

Usage

diag.RBS(
  model,
  mu.link = "identity",
  sigma.link = "identity",
  scheme = "case.weight",
  lx = NULL,
  lz = NULL
)

Arguments

model	Object of class gamlss holding the fitted model.
mu.link	Defines the mu.link, with "identity" link as the default for the mu parameter.
sigma.link	Defines the sigma.link, with "identity" link as the default for the sigma parameter
scheme	Default is "case.weight". But, can be "response", "location" or "precision".
lx	Used in the scheme 'location'.
lz	Used in the scheme 'precision'.

Value

Local influence measures.

Author(s)

Manoel Santos-Neto [e-mail], F.J.A. Cysneiros [e-mail], Victor Leiva [e-mail] and Michelli Barros [e-mail]

References

Leiva, V., Santos-Neto, M., Cysneiros, F.J.A, Barros, M. (2014) Birnbaum-Saunders statistical modelling: a new approach. Statistical Modelling, v. 14, p. 21-48, 2014.

Examples

data(cpd,package='faraway')
attach(cpd)
fit = gamlss::gamlss(actual ~ 0+projected,
family=RBS(mu.link="identity",sigma.link="identity"),method=CG())
Cib <- diag.RBS(fit)$Ci.beta
plot(Cib,ylim=c(0,1),pch=19,ylab=expression(C[i](beta)),xlab="Index")
abline(h=2*mean(Cib),lty=2)
Cia <- diag.RBS(fit)$Ci.alpha
plot(Cia,ylim=c(0,1),pch=19,ylab=expression(C[i](beta)),xlab="Index")
abline(h=2*mean(Cia),lty=2)

santosneto/RBS documentation built on Feb. 5, 2021, 2:12 p.m.